Table 2.
Runtime (in minutes) and memory (in GB) required by ORNA-Q/-K, ORNA, Diginorm and Bignorm for normalizing Brain (147 M) and HeLa dataset (216 M).
| Method | Brain (147 M–35.1 GB) | HeLa (216 M–60.7 GB) | ||||
|---|---|---|---|---|---|---|
| % reduced | time [min] | mem [GB] | % reduced | time [min] | mem [GB] | |
| ORNA | 69.8 | 112 (42) | 6.38 (6.31) | 75.31 | 219 (64) | 9.81 (9.85) |
| ORNA-Q | 70.65 | 116 (50) | 7.10 (7.13) | 72.72 | 223 (70) | 10.01 (10.02) |
| ORNA-K | 70.3 | 130 (52) | 6.41 (6.5) | 73.86 | 279 (75) | 9.98 (10.01) |
| Diginorma | 72.03 | 135 | 12.5 | 72.91 | 198 | 12.51 |
| Diginormb | 70.51 | 112 | 6.26 | 75.10 | 155 | 9.76 |
| *Bignorma | 69.38 | (47) | 41.94 | 71.28 | (58) | 41.94 |
| Bignormb | 69.39 | (41) | 5.23 | 71.26 | (55) | 5.24 |
Notes: The memory required to store the complete dataset in the main memory is indicated in brackets next to the name of the dataset. The column % reduced states the percent of reads reduced by each method. Time and memory as obtained by running the algorithm with 10 threads (if possible) are shown in brackets. *Bignorm always runs with 4 cores and fixed memory settings.